TURNER, HAENKE AND HUBBARD GLACIERS 99 in the Galiano series, early responding to the earthquake shaking. Like the Galiano, it must have advanced within less than two years of the earthquake of September, 1899. This evidence leads us still more strongly to the belief that, in all probability, other small glaciers of this region passed through a cycle of advance prior to 1905, and so long before it that, as in the case of the Miller and Galiano Glaciers, the evidence of the advance was quite hidden by the effect of ablation before the period of our first studies, in 1905. We look upon the discovery of the evidence of this advance of Miller Glacier, the time of which can be so definitely determined, and which falls so appropriately in its place in the series of advancing glaciers, as one of the most striking confirmations of the theory of earthquake cause for advance which has so far been discovered. It is valuable confirmation of a theory when an expected phenomenon, predicted under the theory, can be proved to have occurred. It is further noteworthy, that the advance and transformation of Miller Glacier was less extensive than that of the neighboring Haenke Glacier. This also is expectable under the theory, for with its small size, upon which its early advance depended, a more moderate advance should result than that of the much larger and more slowly responding neighbor. The 1901 photograph does not show the exact end of the newly crevassed glacier so that we cannot now state where it terminated, t Advancing Haenke Glacier. Our evidence of an advance of Haenke Glacier is even more complete than that of Miller Glacier, for, between August, 1905, and June, 1906, it had undergone a complete transformation as was demonstrated by actual observation. The undulating, uncrevassed, moraine-covered surface of 1905 was in 1906 transformed to a perfect labyrinth of crevasses, while clear ice appeared here and there throughout the glacier from as far back in the mountains as we could see down to its front. The Haenke Glacier in 1906 was as badly crevassed as the Atrevida, but there was less clear ice showing, probably because of the slow rate of ablation here where the snow-covered mountains in the background, the ice-covered fiord in front, and Turner and Hubbard Glaciers on either side, give rise to a local climate that is much cooler in summer than in any other part of the coast of Yakutat Bay. This was proved by the presence of snow which covered the alluvial fans in this section in July, the only part of the coast of this inlet where snow remained near sea level. The crevassed surface of the glacier was, in consequence, still so black with debris that even in August we were unable to secure good photographs from a distance. That the broken glacier had also been notably thickened is proved by a comparison of photographs. It can be seen in the earlier photographs (PI. XLI, A) that Haenke Glacier descends over an increased slope just as it reaches the mountain front and th,at its lowermost portion rests on the face of this slope and spreads with decreased slope just beyond it. This condition, which closely resembles that of Turner Glacier, is interpreted, in the same way, as a result of the fact that Haenke valley is a hanging valley in its relation to Disenchantment Bay. In 1906 this increased slope was brought out with far greater clearness partly because the glacier front extended further, but largely because the amount of ice on the slope had been greatly increased. The glacier surface rose much higher, both above the lip of the ice-buried hanging valley and on its frontal slope. At the point where the glacier slope changes in passing over the lip of the hanging valley was found the greatest development of crevassing and the greatest amount of clear ice.